Evaluation of Computer Generated After-Visit Summaries to Support Patient-Centered Care

Following a medical encounter, it is common practice to give patients educational materials and printed instructions in the form of handouts and pamphlets. Many electronic medical records (EMRs) are capable of printing an after-visit summary (AVS) that includes similar content as well as a patient-specific summary of the visit. AVSs have the potential to improve shared decisionmaking, coordination of care, and adherence to recommended plans of care. Content of the summary varies but generally includes the patient’s medications, diagnoses, problems, allergies, and instructions. Little research has been done on how patients use the AVS and whether it is an effective way of conveying information to patients. This project utilized input from patients and clinicians in order to develop three test versions of an AVS, and employed a randomized controlled trial to evaluate the various versions against a standard AVS.

The main objectives of the project were to:

• Ascertain patient attitudes, preferences, and needs regarding the delivery of information at a visit with a primary care physician.
• Identify primary physicians’ attitudes about the utility, content, and value of the AVS.
• Develop and test three different versions of an AVS.
• Disseminate the programming instructions needed to deploy an AVS for health care organizations that use the Epic EMR system.

The project was conducted in four clinics with a diverse payer mix, English and Spanish-speakers, and African-American, Hispanic, and non-Hispanic white clients. The same EMR was used at all four sites, although it was configured differently at each. Interviews were conducted with 12 providers and 48 of their patients in order to inform the content of the three AVS test versions. One of the three versions met Meaningful Use requirements, while the remaining two contained less information.

A total of 68 patients participated in the evaluation phase, with equal numbers of Spanish and English speakers. Two phone interviews were conducted: the first between1-3 days after a visit and the second between 14-21 days after. The purpose of the interviews was to assess the patient’s general AVS content recall, medication recall, satisfaction with the AVS, and self-reported adherence to treatment.

Recall of content of the AVS was low, with the highest being the group who received the shortest AVS version. Although health literacy was lower in the Spanish-speaking population, the study team did not find any significant difference in recall of medications between the groups. Overall, patients were happy with the format of the AVS, and indicated that the medication list and instructions were the most important components. Not surprisingly, Spanish-speaking patients indicated their preference for an AVS in their own language. The mean adherence score did not differ between groups.

The project team concluded that the patients liked to receive an AVS, but the amount and range of the included information did not impact their recall of health-related visit information, or their own perception of the information’s usefulness. However, given the low levels of recall of content it is likely that much of the information on the AVS is not retained by patients. The team concluded that the AVS must be supplemented with efforts to directly remind patients about important information and to not rely solely on the summary.